xNotePadAI — MCP Server

A 13-tool MCP server that gives AI agents persistent, encrypted, searchable memory via notes.

Endpoint: https://www.xnotepadai.com/api/mcp/
Protocol: MCP (Model Context Protocol) JSON-RPC over Streamable HTTP
Auth: Bearer token
Free tier: 1,000 calls/month

Quick Start

{
  "mcpServers": {
    "xnotepad": {
      "type": "http",
      "url": "https://www.xnotepadai.com/api/mcp/",
      "headers": {
        "Authorization": "Bearer YOUR_TOKEN_HERE"
      }
    }
  }
}

Get your token: https://www.xnotepadai.com/settings/ → Your MCP Token (pre-generated, ready to copy)

Tools

Tool	Description
create_note	Create a note (auto-tagged agent-created)
list_notes	List all notes with titles, sizes, dates
get_note	Get full note content by ID
update_note	Update content (agent-created notes only)
delete_note	Delete a note (agent-created only)
search_notes	Semantic search via vector embeddings
ask_notes	Ask a question — AI answers from note content
get_versions	Get version history for a note
tag_note	Add tags to a note
link_notes	Add [[wiki-links]] between notes
merge_notes	Merge two notes (source deleted after)
archive_note	Archive a note
index_notes	Vectorise notes for semantic search

Security Model

• Bearer token auth — SHA-256 hashed, stored in D1, revocable
• Tenant isolation — tenant ID derived from token lookup, never from client args
• Agent sandboxing — agents can only modify/delete notes tagged agent-created
• Rate limiting — 30 req/min per IP
• Zero-knowledge encryption — notes encrypted client-side with AES-256-GCM (MCP accesses unencrypted copies synced by the user)

Use Cases

• AI memory — agents save research, decisions, architecture notes between sessions
• Knowledge retrieval — "What did I decide about auth?" → searches your notes
• Note automation — auto-tag, auto-link, merge duplicates, archive old notes
• Multi-agent workflows — one agent writes, another reads later

Stack

• Cloudflare Workers (runtime)
• Cloudflare D1 (database)
• Cloudflare Vectorize (semantic search)
• Workers AI / Llama 3.3 70B (AI answers)

Docs

• MCP Integration: https://www.xnotepadai.com/mcp/
• API Reference: https://www.xnotepadai.com/developers/
• Privacy: https://www.xnotepadai.com/privacy/

License

MIT